The latest results are promising, with automated transcriptions reaching an average Character Error Rate of 14%. The National Archives enlisted the support of volunteers to produce further training data for their model and they are also looking forward to the possibility of exploiting the collection with the Keyword Spotting tool in Transkribus.